CCIM LogoSandia National Laboratories Logo
HomeCapabilitiesOrganizationAwardsPublications and PresentationsCareer OpportunitiesCSRIPlatforms


2011 Newsnotes

Previous Newsnotes

 

National Research Council (NRC) Committee on Improving the Decision Making Abilities of Small Unit Leaders

The NRC Committee on Improving the Decision Making Abilities of Small Unit Leaders was convened in July of 2010 to examine factors related to decision making among Marine small unit leaders (Company, Platoon, Squad) in the context of counterinsurgency warfare.   Marine small units are increasingly distributed across wide geographic areas and assuming a range of responsibilities, from stabilization and reconstruction missions to pursuing and interdicting insurgents.  To be successful, Marine small unit leaders must be able to interact productively with multiple actors, from local leaders to multinational forces; rapidly assess evolving situations and apply Rules of Engagement to minimize civilian casualties while protecting Marine lives and performing the mission; and do so without direct access to the capabilities and skills available at the battalion level.  The Committee's membership included a range of social, behavioral, physical and operations research scientists from academia, private industry and government.  Laura McNamara, a PMTS in 01465, and retired Sandia Vice President Gerald Yonas were both appointed to Committee, which met with a range of representatives from the Marines, the Office of the Secretary of Defense, and the Navy. Members also observed the Joint Operational Utility Assessment of the Future Immersive Training Environment at Camp Pendleton's Immersive Infantry Trainer, and conducted interviews with recently deployed Marine small unit leaders.  The Committee's findings will be issued in an NRC report in the summer of 2011.

(Contact: Laura McNamara)
February 2011
2011-1647P


Computational Model Reproduces Microcracking in a Crossply Composite

Fiber-reinforced composite materials are known to undergo microcracking due to mechanical loading and changes in environmental conditions. The ability to understand and mitigate microcracking would help materials scientists to develop cheaper, longer-lasting advanced composites. Under the sponsorship of The Boeing Company through the umbrella CRADA, Sandia and Boeing staff are using peridynamics to gain insight into the problem.

A recent success is the simulation by our computation model of important aspects of microcracking in a crossply laminate. In this peridynamic model, carbon fibers are placed randomly within an epoxy matrix, as shown in Figure 1. The model consists of a center layer that contains out-of-plane fibers sandwiched between two outer layers that contain fibers oriented horizontally. The laminate model is stretched until damage appears and progresses. The model predicts that damage starts at random locations within the epoxy matrix in the center layer. As loading increases further, the damaged areas interact and combine with each other to form transverse cracks. The spacing of the cracks is determined by the tendency of the larger cracks to suppress further growth of nearby smaller cracks. The cracks turn as they approach the outer layers of the composite, because the cracks propagate easily between the fibers but not through them.

The model accurately reproduces the important features of cracking in a crossply laminate that are observed in experiments (Figure 2). These features include the spacing of the cracks, which is on the order of the out-of-plane layer thickness, as well as the turning of the cracks at the interface with the outer layers. The model also reproduces the tendency of laminates with very thin center layers to undergo larger global strains before cracking starts than those with thicker center layers. The number of cracks as a function of global strain also agrees qualitatively with experimental data.

The peridynamic model is successful in this application because it provides a consistent nucleation and growth model for cracks within a single mathematical framework. Its computational implementation is meshless, allowing the arbitrary placement of fibers.


Figure 1. Computational model of microcracking in a laminate.



Figure 2. Experimental results on microcracking.

(Contact:Stewart Silling)
January 2011
2011-1374P


IED Prediction

The Joint IED Defeat Organization is funding Sandia to extend their proven Improvised Explosive Device prediction models to provide tactical support to field units in Afghanistan.  After successfully passing an intensive review, the model was identified for further development with the goal of immediate operational deployment. The prediction model involves a combination of machine learning, game theory, and spatial statistics to characterize and predict the cause and effect of various actions by both sides of the conflict. The model was implemented in real-time over four months and successfully predicted IED events weeks in advance with 70-90% percent accuracy in both time and space. The requested extensions will increase the resolution of the model to provide ground and air forces in Afghanistan increased flexibility of in-theater movement and possible response actions, with reduced risk to both coalition forces and local civilian population.

(Contact:David G. Robinson)
January 2011
2011-0862P


DAKOTA Version 5.1 Released

Version 5.1 of the DAKOTA software toolkit was released and deployed in December 2010.  DAKOTA provides mathematical and statistical methods to assist scientists and engineers in iterative analysis studies such as sensitivity analysis, uncertainty quantification (UQ), parameter estimation, and design optimization with computational models.  DAKOTA is used extensively in the Advanced Simulation and Computing (ASC) program at SNL, LANL, and LLNL, in DOE-wide programs such as the Consortium for Advanced Simulation of LWRs (CASL) and Nuclear Energy Advanced Modeling and Simulation (NEAMS), and widely by academic, government, and corporate institutions.

DAKOTA 5.1 includes significant advances in stochastic expansion methods (most notably automated adaptive methods) and extended mixed aleatory/epistemic approaches for UQ.  It has improved statistics reporting as well as variance-based decomposition index calculation and reporting control.  Version 5.1 integrates the new Acro/COLIN optimization framework and Kriging implementation from Surfpack.  In support of this DAKOTA release, the JAGUAR 2.1 user interface will soon be available with improved responsiveness, significant enhancements to its text editor and template management, and a Windows installer.  DAKOTA runs on high-performance computing platforms, as well as on PC and Mac desktop computers.  It is freely available under the GNU LGPL from http://dakota.sandia.gov.  Training classes in its use will be offered at Sandia in the spring of 2011.

(Contact:Brian Adams)
January 2011
2011-0292P



Newsnotes | Info and Events (internal - SNL only) | Open-Source Software Downloads |
Privacy and Security
Sandia National Laboratories Home Page - External or Internal (SNL only)

Maintained by: Bernadette Watts
Modified on: June 10, 2011